Transmission of breast cancer--a controversy resolved.

The Danish breast cancer data collected by Jacobsen (1946) have been reanalysed using morbid risks which incorporate mortality due to breast cancer. A dominant gene is favoured for familial breast cancer, supporting the conclusions of Williams & Anderson (1984) and later authors. Neglect of specific mortality does not greatly alter estimates of gene frequency and displacement, but the evidence for a major gene is inflated. No evidence for heterogeneity was found. Earlier claims of non-Mendelian transmission are in error since we have discovered that transmission probabilities are not correctly implemented in the computer program POINTER. Cases with bilateral breast cancer and males with breast cancer all belonged to families favouring a major gene. Of the cancer sites frequently reported to be associated with familial breast cancer, only ovarian cancer is significant in this material.